LAS VEGAS — Artificial intelligence was the singular obsession at AWS re:Invent this week. And if there were any doubts about the outsized ambitions of parent company Amazon.com Inc., those were dispelled by its bet-the-house gambit.
To use a Vegas analogy, Amazon placed chips on every number and color at the roulette wheel in its pursuit of world domination of tech’s biggest revolution yet. AWS unleashed a wave of AI news, spotlighted guest stars from Big Tech to vouch for AWS’s new AI products, and shared a blueprint of investments in generative AI startups that will extend its tentacles into nearly every facet of life.
“AWS’s AI-first strategy was on full display during [CEO] Matt Garman’s keynote [on Tuesday], particularly with the introduction of groundbreaking multi-agent capabilities on Amazon Bedrock,” said Chris Jangareddy, AWS Data & AI Lead at Deloitte Consulting LLP. “By enabling orchestration agents to manage complex, multi-step workflows, AWS is redefining how businesses can design dynamic, intelligent systems tailored to their unique needs. This innovation reflects a maturing AI landscape where collaboration between AI agents can drive new efficiencies and insights.”
The onslaught continued on Wednesday, when AWS paraded the first curated set of AI apps that can be fully managed and secured by SageMaker. AWS also unfurled SageMaker Unified Studio, a single spot to find and work with data from across an organization. And NVIDIA Corp. said its NVIDIA NIM microservices are now available on AWS to support optimized inference and lower latency for genAI app deployments.
“The next generation of SageMaker brings together capabilities to give customers all the tools they need for data processing, machine learning model development and training and generative AI, directly within SageMaker,” Swami Sivasubramanian, vice president of data and machine learning at AWS, said in a keynote speech on Wednesday.
Earlier this week, former AWS CEO Andy Jassy (who is now Amazon CEO) made his first stage appearance in three years at the show to introduce Nova, a new family of foundation models that lend support for processing text, images and video in a range of multimodal tasks. Garman discussed a new generation of SageMaker, including an AI version, that brings together fast SQL analytics, petabyte-scale big data processing, data exploration and integration, model development and training, and generative artificial intelligence into one integrated platform.
To combat hallucinations, AWS unveiled Automated Reasoning checks. It validates an AI model’s responses by cross-referencing customer info for accuracy. Not to appear laggard, AWS also previewed a cloud service based on its Tranium2 processors.
In a high-profile endorsement of Amazon’s status in the AI race against cloud rivals Alphabet Inc. and Microsoft Corp., Benoit Dupin, Apple Inc.’s senior director of machine learning and AI, made a surprise cameo during Tuesday’s keynote address, to outline the iPhone maker’s use of Amazon services for products and services, as well as AWS’s behind-the-scenes role in Apple Intelligence. He said Apple used AWS chips Amazon Graviton and Inferentia to achieve gains in efficiency of more than 40%.
Apple is also in the early stages of evaluating Trainium2, according to Dupin.
Additionally, AWS executives shared a list of investments via accelerator programs aimed to give Amazon a leg up in battery efficiency (FlexGen), autonomous buildings (BrainBox AI), de-carbonization of data centers (Orbital Materials), and clean energy (Thea Energy).
“We are on the verge of so many things that will improve with GenAI — health care and life sciences will come up with cures to illnesses, new forms of alternative energy, you name it,” Kathryn Van Nuys, head of startup business development at AWS, said in an interview. One of the AI companies that Amazon initially backed, Anthropic, now boasts an $8 billion investment from Amazon.
“These startups have some of the greatest minds in technology,” said Lisbeth Kaufman, head of climate tech at AWS.
It all adds up to a tidal wave of innovation, not to mention market dominance, by Amazon and AWS in the months and years to come as it battles to play catchup with Microsoft and Google in the AI race. As it has before, Amazon is aggressively branching into a myriad of industries that will eventually pay off down the road, say analysts.
“Amazon Nova highlights AWS’s commitment to delivering unprecedented flexibility and choice in AI. By offering a diverse suite of foundation models, AWS empowers organizations to select the tools best suited for their specific workloads and industries,” JB McGinnis, AWS Lead at Deloitte Consulting LLP, said. “This approach realizes that no two businesses are the same, and ensures that companies can tailor AI solutions to their unique needs. This exemplifies the shift towards adaptable AI systems, giving enterprises the ability to innovate with precision and confidence while maintaining control over costs and scalability. It’s a critical step in enabling broader adoption across industries.”
Said Wedbush Securities analyst Dan Ives, summarizing the conference: “We are encouraged by the pace of innovation at AWS and come away from re:Invent with increased conviction in the trajectory of AWS growth and the health of Amazon’s competitive positioning relative to peers.”
